PDA

View Full Version : Quantum dot CCDs?



EDG
2010-Mar-23, 04:11 AM
http://news.bbc.co.uk/2/hi/technology/8580372.stm

Sounds like this might revolutionise CCD cameras if it takes off? (if I'm reading this right anyway).

RickJ
2010-Mar-23, 05:05 AM
I see a red flag or two here. They claim today's CCD's are only about 25% efficient. Many astro CCD's approach 80 and even 90% efficiency. Then there's that use of the "magic word" quantum. That usually means; look out what follows is overblown hype or worse. Usually worse.

While shrinking the size of a pixel can mean a smaller camera for a cell phone or spy it is often meaningless in astronomy. For instance my camera uses 9 micron pixels but the atmosphere so blurs things I can't begin to use that except on very rare nights. Anything smaller would be a worthless to me.

A smaller chip size is the exact opposite of where astro imaging CCD's is going. Back in my film days many used medium format size film and lenses to match. This is an order of magnitude larger than any CCD chip made though there's a push toward larger and larger chips. The reason is they can take in far more sky. Right now many objects requires a mosaic of many images to take in the object, a smaller chip would make this worse not better. That would increase exposure time not decrease it!

If I were making a spy camera then this might be good IF the hype is true, which I doubt. If it makes for cheaper chips that might help. What is the chip's noise factor, dark current, spectral response, well depth for just a few important but unanswered questions.

I'm far from convinced. I've seen such hype before that went nowhere.

Rick

EDG
2010-Mar-23, 05:42 AM
Why wouldn't they just be able to make a bigger version of this though? I'm not seeing anything that says it has to be in a small camera - if you made a normal sized CCD with this quantum dot material, then it seems you could get a lot MORE light than you would get with the existing technology.

RickJ
2010-Mar-23, 08:24 AM
As I mentioned, today's chips are up to 90% efficient. They are claiming 25% which isn't true (it is true of silicon solar cells however). So with only 10% more to go that's not significant. The whole article is about reducing chip size by reducing pixel size for smaller cameras. Nothing about increasing chip size. Smaller pixels would be useless for most imagers creating severe over sampling and an increase in exposure time, not decrease.

Quantum dot is a solar cell technology that may hold some promise. They are nano particles with optical characteristics. But solar cells are very different from an imaging chip and I've seen nothing indicating they could be used this way. Or if they were how this would be an advantage as they seem to increase efficiency in a random way. Not a problem when generating electricity but when a highly accurate photon count is the goal this could really make for a noisy image compared to today's CCD's. They also tend to be very frequency tuned by controlling the size of the dot. That would mean bigger dots for red than blue thus likely differing resolution for red and blue. Not good for scientific imaging. If this could be harnessed in same size pixels it would mean an end to color filters in OSC cameras which would create an efficiency gain but it wouldn't be all that great, 10 or 15%.

Until the questions about other relevant factors I mentioned earlier are answered as well as the ability to use it in far larger pixel sensors of large area I'm not going to follow the hype.

Rick

TrAI
2010-Mar-30, 11:14 AM
As I mentioned, today's chips are up to 90% efficient. They are claiming 25% which isn't true (it is true of silicon solar cells however). So with only 10% more to go that's not significant. The whole article is about reducing chip size by reducing pixel size for smaller cameras. Nothing about increasing chip size. Smaller pixels would be useless for most imagers creating severe over sampling and an increase in exposure time, not decrease.

Quantum dot is a solar cell technology that may hold some promise. They are nano particles with optical characteristics. But solar cells are very different from an imaging chip and I've seen nothing indicating they could be used this way. Or if they were how this would be an advantage as they seem to increase efficiency in a random way. Not a problem when generating electricity but when a highly accurate photon count is the goal this could really make for a noisy image compared to today's CCD's. They also tend to be very frequency tuned by controlling the size of the dot. That would mean bigger dots for red than blue thus likely differing resolution for red and blue. Not good for scientific imaging. If this could be harnessed in same size pixels it would mean an end to color filters in OSC cameras which would create an efficiency gain but it wouldn't be all that great, 10 or 15%.

Until the questions about other relevant factors I mentioned earlier are answered as well as the ability to use it in far larger pixel sensors of large area I'm not going to follow the hype.

Rick

I believe the main reason CCDs have low efficencies in consumer cameras are the Bayer mosaic filters neccesary for color reproduction, high sensitivity astrophotography sensors do not have these filters.

But how quantum dots would help, I am less sure of, I suppose the most obvious way would be a multilayer sensor approach, if the dots will let for example, shorter wavelengths through, one might have the red sensitive dots in the upper layer, then green, and then blue. Or some other combination, depending on what makes the most efficent setup.

As for pixel size, I would assume that more blue dots would fit than red dots on the same area, but perhaps more red photons will have a probability of hitting each individual red dot than a blue photon has of hitting a blue dot, so having more blue dots would be a good thing.

Noise may be an issue, though it depends on how it compares to the noise from current sensors, the sensors in phones and cheap, small cameras are not exacly free of noise, you know.

In astrophotography the selectivity of frequency in the dots may be a problem, I suppose, though it may work in applications where color selectivity is needed, for example space craft cameras often use color wheels with several different wavelengths, if the CCDs could be layered so that each layer coresponds to the frequencies that are of interest, it may replace the wheel, and allow taking pictures in many frequencies for identifing materials. And such sensors would probably be cooled to reduce noise.

kris27
2010-Apr-01, 12:42 AM
Hm interesting. But the size of the pixel is quite crucial in light gathering. The smaller pixel than less photons are captured (area is small) so need to extend exposure time. I thing that this is just hype ....

TrAI
2010-Apr-01, 02:07 AM
Hm interesting. But the size of the pixel is quite crucial in light gathering. The smaller pixel than less photons are captured (area is small) so need to extend exposure time. I thing that this is just hype ....

I think the idea was that small CCDs for compact devices would be more sensitive if based on quantum dots instead of the traditional design for the same area per pixel, not that the pixels would necessarily be smaller.